Pentax/Camera Marketplace |
Pentax Items for Sale |
Wanted Pentax Items |
Pentax Deals |
Deal Finder & Price Alerts |
Price Watch Forum |
My Marketplace Activity |
List a New Item |
Get seller access! |
Pentax Stores |
Pentax Retailer Map |
Pentax Photos |
Sample Photo Search |
Recent Photo Mosaic |
Today's Photos |
Free Photo Storage |
Member Photo Albums |
User Photo Gallery |
Exclusive Gallery |
Photo Community |
Photo Sharing Forum |
Critique Forum |
Official Photo Contests |
World Pentax Day Gallery |
World Pentax Day Photo Map |
Pentax Resources |
Articles and Tutorials |
Member-Submitted Articles |
Recommended Gear |
Firmware Update Guide |
Firmware Updates |
Pentax News |
Pentax Lens Databases |
Pentax Lens Reviews |
Pentax Lens Search |
Third-Party Lens Reviews |
Lens Compatibility |
Pentax Serial Number Database |
In-Depth Reviews |
SLR Lens Forum |
Sample Photo Archive |
Forum Discussions |
New Posts |
Today's Threads |
Photo Threads |
Recent Photo Mosaic |
Recent Updates |
Today's Photos |
Quick Searches |
Unanswered Threads |
Recently Liked Posts |
Forum RSS Feed |
Go to Page... |
|
58 Likes | Search this Thread |
03-04-2021, 10:40 AM | #91 |
Staff note: This post may contain affiliate links, which means Pentax Forums may earn a small commission if a visitor clicks through and makes a purchase. If you would like to support the forum directly, you may also make a donation here. You simply cannot make these kinds of decisions from the size of the sensor or the number of MP. The whole image producing process, from AD conversion to internal noise reduction counts. You can't clean up an image if major artifacts are produced in the image creation process, like false colour and moire, no matter how much you reduce the image. The original data is compromised. There's nothing to work with. And I already know, you don't believe in standardized tests, only your opinion counts. Please don't bother with that. But, in your opinion, why is a 50 MP camera rated lower than a 36 MP image, when noise is standardized? (because that's what Dxo does that no one else does. It standardizes noise in thier tests.) Last edited by normhead; 03-04-2021 at 01:51 PM. | |
03-04-2021, 04:06 PM | #92 |
Then why is a 5Ds rated so much lower when DxO standardizes the noise reduction, than a K-1? 10 DxO points less? These things are never universally true. So, yes, I would investigate, with real images. The way you investigate is side by side comparisons of images from both cameras. Any assumption has a chance of being wrong. You simply cannot make these kinds of decisions from the size of the sensor or the number of MP. The whole image producing process, from AD conversion to internal noise reduction counts. You can't clean up an image if major artifacts are produced in the image creation process, like false colour and moire, no matter how much you reduce the image. The original data is compromised. There's nothing to work with. And I already know, you don't believe in standardized tests, only your opinion counts. Please don't bother with that. But, in your opinion, why is a 50 MP camera rated lower than a 36 MP image, when noise is standardized? (because that's what Dxo does that no one else does. It standardizes noise in thier tests.) The lure of a 61mp sensor pixel shifting to produce a 240mp image (which I'd probably do just once for kicks) must be resisted. | |
03-04-2021, 11:01 PM | #93 |
Then why is a 5Ds rated so much lower when DxO standardizes the noise reduction, than a K-1? 10 DxO points less? These things are never universally true. So, yes, I would investigate, with real images. The way you investigate is side by side comparisons of images from both cameras. Any assumption has a chance of being wrong. You simply cannot make these kinds of decisions from the size of the sensor or the number of MP. The whole image producing process, from AD conversion to internal noise reduction counts. You can't clean up an image if major artifacts are produced in the image creation process, like false colour and moire, no matter how much you reduce the image. The original data is compromised. There's nothing to work with. And I already know, you don't believe in standardized tests, only your opinion counts. Please don't bother with that. But, in your opinion, why is a 50 MP camera rated lower than a 36 MP image, when noise is standardized? (because that's what Dxo does that no one else does. It standardizes noise in thier tests.) In practice, there are not many who will use more than a few times 6400 ISO on cameras like 5DsR, which was released for studio photographers and it wasn't optimized for high ISO at that time (6 years ago), so it's a less chance to get major artifacts at this ISO. For those few occasions when you shoot at ISO 6400, you have the resize option to make the image look with less noise. Since you promote Imaging Resource as your trusty source when comes to cameras and lenses, you can see in their article that even at ISO 6400 you can get a nice big print from A7R IV. I quote from Imaging Resource "ISO 6400 prints begin to show some detail loss, particularly in our fabric swatches, but detail elsewhere in our test image, particularly higher contrast areas, are still pleasing and sharp. Noise is definitely noticeable in the shadow areas. Again, the A7R IV can print a size better at this ISO than the former model, allowing for a very good 16 x 20 inch print. A 20 x 30 might be possible with careful image editing and processing." If a 16 x 20 inch print look very good at 6400 ISO, then a resized image from the same camera will look even better in terms of noise once it's resized from 61mp to 20mp for viewing it on screen. If you want to make a point with ISO 16000 or even more where some artifacts may become a problem, go ahead. I don't care much about artifacts at ISO 16000 or more because I wouldn't buy a Sony A7R IV to shoot at such ISO. And I don't shoot things so that in post I have to push more than 2 stops the shadows because I'm careful with the exposure as much as I can and seeing the image in viewfinder before pressing the shutter also helped in this regard. I would resize instead a ISO 6400 image from 61mp to 30mp to gain a stop in terms of noise if I would have to send the images to a client. Because Normhead, I send images to clients, not test charts, and if this method helped all the pro photographers who shared this tip to me, it sure helps me also. Here is the link of the Imaging Resource article. Sony A7R IV Review - Image Quality Also, at comments no. 68 and 84 I pointed out how light and small a full frame can be with only 3 lenses that can cover the focal lenghts from 17mm to 180mm, at a fast f2.8 aperture. Or how light and small an APS-C can be with only 2 f2.8 lenses (Tamron 17-70mm and Tamron 70-180mm). And guess what, the full frame combo is 2000 grams easier than K1 with equivalent lenses and it's even 40$ cheaper. That's the power of third party support that makes Sony more appealing than Canon and Nikon mirrorless for the ones on a budget or for the ones who want a light combo with high performance. Last edited by Dan Rentea; 03-05-2021 at 07:11 AM. | |
03-05-2021, 01:03 AM | #94 |
If the difference in image quality when pixel peeping (I take that as '@ 100%') is negligible, then surely having more megapixels ("with no discernible difference in quality") is beneficial should one need apply heavier crops. I will agree, however, that the use of accelerator chips has indeed helped the Pentax offerings of late. The KP is impressive at higher ISOs and the K3iii seems to be excellent in that regard also (albeit yet to reach market). | |
03-05-2021, 02:43 AM - 2 Likes | #95 |
The DxO list you posted elsewhere (here), with reference to the Pentax 645Z and K-1 being in #2 and #11 points respectively (although I counted the K-1 at #12) glossed over #3-#8 and #10-11 being higher than 36MP and rated higher than the K-1. That list includes the A7R IV at 61MP in #6. The reference to a specific 50MP camera being lower than the K1 is reasonably moot given current offerings. If the difference in image quality when pixel peeping (I take that as '@ 100%') is negligible, then surely having more megapixels ("with no discernible difference in quality") is beneficial should one need apply heavier crops. I will agree, however, that the use of accelerator chips has indeed helped the Pentax offerings of late. The KP is impressive at higher ISOs and the K3iii seems to be excellent in that regard also (albeit yet to reach market). DXO results for Pentax K1: Portrait (Color Depth): 25.4 bits Landscape (Dynamic Range): 14.6 Evs Sports (Low-Light ISO): 3280 ISO Overall Score: 96 DXO results for Sony A7R IV: Portrait (Color Depth): 26 bits Landscape (Dynamic Range): 14.8 Evs Sports (Low-Light ISO): 3344 ISO Overall Score: 99 I already posted the total price of K1 II with the trinity of lenses vs. Sony A7R IV with the Tamron new trinity of lenses and Sony was 45$ cheaper and 2070g lighter while having basically 2 cameras in one: a 61mp full frame and a 26mp APS-C camera which according to DXO will take better high ISO images than K1. But, there are many reasons I don't look at DXO or lab tests and the most important one is the fact that I'm interested in what a camera can do in real life situations, in the day by day shooting conditions where I normally use the camera. DXO tests in controlled environment, with camera on a tripod, means absolutely nothing to me. A few weeks ago I was the godfather at a christening. I gave my camera to my nephew and the ISO was set to Auto ISO with the upper limit of 12800 ISO. The official photographer had Canon R6 which among other things is very good at high ISO. I saw the differences when we received the full resolution images compared to my EOS R in terms of noise performance at same ISO. That being said, resized both of the images taken with R6 and EOS R to 2048px on the long edge for normal view on social media or websites, the difference in terms of noise was gone, which is why I said that resizing an image helps when comes to overall noise. Here is an example with me being photographed by my almost 12 years old nephew and by the official photographer. The image taken by my nephew was at ISO 12800 because he had exposure compensation on +1 also and I was a little against the direction of the light. I also cropped the head from both images at 100% so that you can see there is a noticeable difference in terms of noise if the files are seen at full size resolution and zoomed in. I would have no problem sending the 12800 ISO image from this example to a client, resized from 6700px which is the normal file size of EOS R to 4096px (4k resolution) because at 4k resolution the clients won't tell the difference between my image taken at 12800 ISO and the image taken by the official photographer at 3200 ISO, unless they pixel peep each and every image. I wouldn't trade my EOS R with its DXO Sports (Low-Light ISO) score of 2742 ISO for R6 with its (Low-Light ISO) score of 3394 ISO because resizing my files from 30mp to 20mp close the gap between the two cameras in terms of noise. I would trade my EOS R for R6 for other reasons, eye af being the most important of them. The noise on the head can be better viewed by clicking the Download the original image button. The zooming button on the link messes up the images. Last edited by Dan Rentea; 03-05-2021 at 07:17 AM. | |
These users Like Dan Rentea's post: |
03-05-2021, 02:53 AM - 2 Likes | #96 |
If I take the DXO test as the absolute and incontestable truth, then I would care even less about the noise on A7R IV because according to DXO, A7R IV has a better noise performance than Pentax K1, with extra 25 more mp on Sony camera. DXO results for Pentax K1: Portrait (Color Depth): 25.4 bits Landscape (Dynamic Range): 14.6 Evs Sports (Low-Light ISO): 3280 ISO Overall Score: 96 DXO results for Sony A7R IV: Portrait (Color Depth): 26 bits Landscape (Dynamic Range): 14.8 Evs Sports (Low-Light ISO): 3344 ISO Overall Score: 99 I already posted the total price of K1 II with the trinity of lenses vs. Sony A7R IV with the Tamron new trinity of lenses and Sony was 45$ cheaper and 2070g lighter while having basically 2 cameras in one: a 61mp full frame and a 26mp APS-C camera which according to DXO will take better high ISO images than K1. But, there are many reasons I don't look at DXO or lab tests and the most important one is the fact that I'm interested in what a camera can do in real life situations, in the day by day shooting conditions where I normally use the camera. A few weeks ago I was the godfather at a christening. I gave my camera to my nephew and the ISO was set to Auto ISO with the upper limit of 12800 ISO. The official photographer had Canon R6 which among other things is very good at high ISO. I saw the differences when we received the full resolution images compared to my EOS R in terms of noise performance at same ISO. That being said, resized both of the images taken with R6 and EOS R to 2048px on the long edge for normal view on social media or websites, the difference in terms of noise was gone, which is why I said that resizing an image helps when comes to overall noise. Here is an example with me photographed by my almost 12 years old nephew and by the official photographer. The image taken by my nephew was at ISO 12800 because he had exposure compensation on +1 also and I was a little against the direction of the light. I also cropped the head from both images at 100% so that you can see there is a noticeable difference in terms of noise if the files are at full size resolution. I would have no problem sending the 12800 ISO image from this example to a client, resized from 6700px which is the normal file size of EOS R to 4096px (4k resolution) because at 4k resolution the clients won't tell the difference between my image taken at 12800 ISO and the image taken by the official photographer at 3200 ISO. I wouldn't trade my EOS R with its DXO Sports (Low-Light ISO) score of 2742 ISO for R6 with its (Low-Light ISO) score of 3394 ISO because resizing my files from 30mp to 20mp close the gap between the two cameras. I would trade my EOS R for R6 for other reasons, eye af being the most important of them. Good breakdown. This is why the A7iii has won out for me. It's small and light, especially with native lenses, for a full-frame offering and 24mp is a perfect spot for an amateur photographer such as myself to take high ISO pictures without being overly concerned about noise with minimal effort. If I wanted to go about resizing photos and suchlike, the A7Riii would probably win out for me (and, truthfully, is tempting just because of additional cropability). It really all does come down to the user and what benefits them personally. I've just decided in this instance to play devil's advocate by highlighting that there are high-MP full-frame offerings that are close to top of the leader board. The Tamron trinity + Sony is certainly a tempting prospect for many. I've recently joined a couple 'Sony A7xxx users' groups on Facebook and the numbers using the 28-75mm f/2.8 is astonishing. There's an increasing trend towards new purchases of the 17-28mm f/2.8 too which, from my sporadic research, appears to be an outstanding lens. | |
These users Like Benz3ne's post: |
03-05-2021, 07:25 AM | #97 |
There's Dan, two pages of distraction, the for the most part agrees with me. But little of it affects the original criticism, accept for the part where he admits in a round about way that I'm right, because he had no alternative. Quote: 5DsR, which wasn't optimized for high ISO Exactly my point. You can't tell from sensor size how the camera will handle noise, artifacts etc. How the internal processing is configured means more. As for reducing image sizes, everyone does that. You don't have to go to a conference to find out that kind of silliness. What you haven't shown is that a 50 MP image reduced will be better than a 36 MP image reduced to say 16 MP, or even to 36 MP. When you consider that because of artifacts, cross talk etc the 50 MP image was less useable pixel peeping the 50 MP image actually has to be cleaned up quite a lot just to catch up with the larger pixel camera, if it can ever completely match it, is open to debate, not whether or not it's better. In the case of your example, it is pretty much equal. No meaningful advantage to the hi MP camera is demonstrated. Given that an A7R IV costs $1000-$1500 more that a K-1, I'd venture that anyone but an action photographer or possibly a wedding photographer would probably wasting their money on unless they are the proverbial 145 pound weakling. The forum consensus from back when these were discussed was that 3 DxO points was pretty much indistinguishable, that 5 was the minimum needed for even a pixel peaking difference. So from that analysis, the three point difference is pretty much meaningless, and you've once again proved my point not yours. I'm comparing to camera I actually use. Cherry picking the the A7R IV also proves my point. You had to find a camera on the list, not just any camera, not just any full frame. You cherry picked data. Your lack of experience in interpreting results kind of means you mislead yourself and others. Just an example.. So lets look at the numbers you posted then crowed about for high ISO, 3340 compared to 3280. A difference of 160 ISO. That's a difference of less than 2% even if you ignore the exponential nature of ISO measurement. Even smaller if you don't. As I said, indestinguishable. In photographic terms, talking less than1/3 stop is silly. the next stop after 3200 is 6400. What you're talking about is less than .05% of a stop. So essentially, the numbers in your post made my point. The extra resolution gets you next to nothing. Your interpretation of those numbers however is seriously flawed.They support my point, not yours. According to the DxO numbers there is no advantage to image quality using an A7R over a K-1 despite the infinitesimally small numerical advantage. Thats why direct comparison and empirical evidence is so important when making comparisons. You can point out a "difference" using numbers that don't exist unless you're using test equipment. You can know the numbers, but as in this case, if you don't know how to correctly interpret them, you'll come to erroneous conclusions. Your problem here being, you don't seem to understand the difference between a numerical difference, and a significant numerical difference. Every valid scientific paper will give you both. The raw data, and the standard deviation. Bottom line, you haven't done what you need to do, to make the point you're trying to make. And the evidence you present can be interpreted both ways. You dwell on infinitesimally small differences to try and claim a significant difference, and anyone familiar with interpreting science (or stats, or math) will tell you how insignificant those difference are. Bottom line, the K-1 performs at almost exactly the same level as the A7R IV by any meaningful evaluation, and if that's the best example you have, you've completely lost your case, trying to prove the value of higher res. You proved my point, not yours. Well, at least to anyone who knows how to interpret the data. I have no doubt the uneducated will be impressed by your rambling 2 page answer to a two line question, most of it irrelevant to the topic. They always are. The fact that other equally statistically challenged people might agree with your point doesn't make it the truth. It just means a lot of people haven't got a clue, and you like to hang with them and repeat their fallacies. If your point was reducing image size can clean up an otherwise unacceptable image, everyone knows that. I have lenses I will use only if I know the final image will be reduced, because at full size pixel peeping they are soft, but reduced in size they actually poll better than some sharper lenses if you run a poll asking which image people prefer. (I'm talking about the FA 35-80 and FA-J 18-35 here.) That reducing images size on an A7R IV makes it preferable to a K-1 image isn't supported by this data. You have ably demonstrated here why a hypothesis is just hypothetical, until there is empirical evidence to back it up, and how someone can make erroneous conclusions by not understanding the meaning of the raw stats. If you could see the actual images DxO used to derive their numbers, you wouldn't have made these mistakes. You would have seen the images are visually identical. And you will continue to make such mistakes until you start doing direct comparisons, or take a stats course. Just learn what a significant difference is, and why it's important in numerical evaluations. Here, check this out. A Refresher on Statistical Significance Last edited by normhead; 03-05-2021 at 10:20 AM. | |
03-05-2021, 11:29 AM - 2 Likes | #98 |
As for reducing image sizes, everyone does that. You don't have to go to a conference to find out that kind of silliness. What you haven't shown is that a 50 MP image reduced will be better than a 36 MP image reduced to say 16 MP, or even to 36 MP. when you consider that because of artifacts, cross talk etc the 51 MP image was less useable pixel peeping the 50 MP image actually has to be cleaned up quite a lot just to catch up with the larger pixel camera, if it can ever completely match it, is open to debate, not whether or not it's better. IN he case of your example. it is pretty much equal. No meaningful advantage to the hi MP camera is demonstrated. If you have images taken side by side from A7R IV and lower resolution cameras like K1 or D810 for example, taken in real life conditions, we can take a look and draw some conclusions. I was able to provide an example with a just released camera with 20mp that has a much better performance than a 30mp EOS R camera, at high ISO. The forum consensus from back when these were discussed was that 3 DxO points was pretty much indistinguishable, that 5 was the minimum needed for even a pixel peaking difference. So from that analysis, the three point difference is pretty much meaningless, and you've once again proved my point not yours. Just an example.. So lets look at the numbers you posted then crowed about for high ISO, 3340 compared to 3280. A difference of 160 ISO. That's a difference of less than 2% even if you ignore the exponential nature of ISO measurement. Even smaller if you don't. As I said, indestinguishable. In photographic terms, talking less than1/3 stop is silly. the next stop after 3200 is 6400. What you're talking about is less than .05% of a stop. So essentially, the numbers in your post made my point. The extra resolution gets you next to nothing. Your interpretation of those numbers however is seriously flawed.They support my point, not yours. According to the DxO numbers there is no advantage to image quality using an A7R over a K-1 despite the infinitesimally small numerical advantage. Thats why direct comparison and empirical evidence is so important when making comparisons. You can point out a "difference" using numbers that don't exist unless you're using test equipment. You can know the numbers, but as in this case, if you don't know how to correctly interpret them, you'll come to erroneous conclusions. Bottom line, the K-1 performs at almost exactly the same level as the A7R IV by any meaningful evaluation, and if that's the best example you have, you've completely lost your case, trying to prove the value of higher res. You proved my point, not yours. Well, at least to anyone who knows how to interpret the data. I have no doubt the uneducated will be impressed by your rambling 2 page answer to a two line question, most of it irrelevant to the topic. They always are. I strongly suggest you to stop offending people you know nothing about. You don't know who I hang out with and therefore I expect apologies here. You can contradict me on a subject without offending me or my friends by telling they don't have a clue about photography. Last edited by Dan Rentea; 03-05-2021 at 04:16 PM. | |
These users Like Dan Rentea's post: |
03-06-2021, 11:10 PM - 2 Likes | #99 |
Hi folks. Before tempers force the moderation team to take the well known actions, please remain civil, and respectfully agree to disagree. Thanks. | |
These users Like Unregistered User 8's post: |
03-08-2021, 12:01 PM | #100 |
Bear in mind I’m not a grip person. I own three, none are regularly used. I got it because it was ridiculously cheap and occasionally I’ve used one with the cameras that I have with one (OMD-EM1, K-3). The fit and finish are very good. The controls work as intended. The remote that came with the grip looks nice but I haven’t used it yet. I have never used the Sony branded grip, so I can’t provide any clear detail on comparison. I don’t like the Sony method of attaching a grip, having to remove the battery door vs having dedicated contacts seems goofy. The grip does have dedicated storage for the door but it still isn’t as easy as the k-3/em1 where the grip just goes on. On the other hand, having both batteries accessible from the grip directly for replacement/external charging is easier than having to remove the grip to access a battery that’s depleted. SmallRig L-Bracket for Sony A7R IV LCS2417 - SmallRig SmallRig A7R4 Camera L Plate L Bracket for Sony A7R IV W/ Arca compatible base plate & side plate 2417|Camera Cage| - AliExpress SMALLRIG L-Bracket L Plate for Sony A7R IV - LCS2417: Amazon.co.uk: Camera & Photo Gives a left hand hold as well as a bit more protection too. Last edited by Parry; 03-08-2021 at 02:37 PM. | |
|
Bookmarks |
Tags - Make this thread easier to find by adding keywords to it! |
135mm, 2x, 50mm, angle, aps-c, camera, crop, f/1.8, f/2.8, fe, gain, landscape, lens, lenses, light, loss, mode, mp, noise, praise, product, purchase, purposes, range, reviews, sensor, sony, tamron, teleconverter, zeiss |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
FF lens with limited budget - of legacy | mbukal | Pentax SLR Lens Discussion | 19 | 12-30-2019 11:36 AM |
My budget for FF.......... | RonHendriks1966 | Pentax Full Frame | 107 | 02-11-2016 07:55 AM |
Pentax FF setup versus APSC budget | biz-engineer | Pentax Full Frame | 89 | 10-02-2015 10:48 AM |
Low budget FF digital? | ihasa | Canon, Nikon, Sony, and Other Camera Brands | 28 | 06-17-2012 07:52 PM |
Best prime lens trio with upgrading possibility to FF | MichaelBilson | Pentax SLR Lens Discussion | 29 | 05-16-2012 03:17 PM |