[Lew Dite]

This article from 2015 recommends some cameras for astrophotography. The authors mention the K-3II but admit they never used one. Concerning it they say,

...at high ISOs noise levels are slightly worse than with the [Canon] 70D, and certainly worse than the [Nikon] D7200. Noise is on par with the Canon D750/760 (sic) models. OK but not the best.

They don't reveal what data they looked at to arrive at those conclusions. They also don't say what kind of noise they are talking about. It can't have been read noise since the published data I have seen on that says that the K-3 and K-3II are considerably better than the 70D and 80D (the 70D's successor) at almost every ISO value.

Nor do their conclusions match the data at the Sensor Noise DB, which is the only source of data I have seen on thermal noise during long exposures (30 seconds and greater). That data says that the D7200 has worse thermal noise than either the 70D or K-3 (no data on K-3II in the DB), which contradicts one of the statements quoted above. However, Roger Clark claims here that Brendan's measurement methodology is seriously flawed and can't be trusted.

So, some questions:　

(1) Is there a source of reliable data on thermal noise of cameras over long exposures (aside from the questionable Sensor Noise DB)?

(2) Which camera models are comparable to the K-3II in terms of thermal noise over long exposures?

Answering (2) without resorting to opinion would seem to require that there be an answer other than "no" for (1).

[maw]

Thermal noise during long exposures: K-3II Vs. others

I have a test on four DSLRs in a magazine and it seems to me the opposite or maybe very close, don't let you influence. I've prepared the test between pentax K3 and Canon 70D,

good reading.

Nikon vs Pentax vs Canon

[jbinpg]

I have used K-5, K-3 and K-1 for astro. As with all Pentax Exmor sensors, they are excellent for astro. Don't believe everything you read.

[k5astro]

I haven't seen any thermal noise data but for sensor data I go to DXO for colour info and Sensorgen - digital camera sensor data for noise and DR info.

[beholder3]

The Sensor Noise dB is the only Thinge you Need.

Its results are shown as Actual images which is much better than some kids doing funny calculations.

The proof is in the Image.

Roger Clark seems quite the Self important theoreticist, without substantial Arguments and certainly without any photos to show better comparisons.

[Lew Dite]

I have a test on four DSLRs in a magazine and it seems to me the opposite or maybe very close, don't let you influence. I've prepared the test between pentax K3 and Canon 70D,

good reading.

Nikon vs Pentax vs Canon

Thanks for the link. I was hoping it was to an online version of that test you mention. But it is only to DxOMark's comparison of the three cameras, which I have read. Unfortunately, none of DxOMark's data concerns long-exposure thermal noise.

(Did you do the test/comparison you refer to or by "I have" do you mean you simply have a copy of the magazine with the test in it? Regardless, can you name the magazine and the issue? I would like to read that article.)

[MossyRocks]

I wouldn't worry about thermal noise from a K-3II for astrophotography. That is the camera that Pete_XL uses and gets some phenomenal results. Once you start getting some idea of what you are doing I would suggest taking some dark frames (lens cap on same ISO and shutter speed as the rest of the images) and some bias frames (lens cap on same ISO as the rest of the images but shutter is as fast as it will go) and then load them into deep sky stacker when you start your processing. Doing that should help with any noise issues that may be of concern.

For an example of what one can do I like to point people to this thread started by Pete_XL with a picture of the orion nebula taken with his K-3II using a Pentax SMC M 200mm lens.

[Lew Dite]

Roger Clark seems quite the Self important theoreticist, without substantial Arguments and certainly without any photos to show better comparisons.

I don't believe in discounting the message because of the messenger. If any of Roger's statements have validity then they have validity independent of his personality. By the same token any of Roger's points that don't have validity won't become valid just because he says they are. Science, right? In any event, many of Roger's photos looks quite good in my opinion. I wouldn't deign to criticize them until I have achieved better. (Even then why would I do that? This isn't a contest and I have better things to do with my time.)

---------- Post added 04-23-18 at 06:45 AM ----------

I wouldn't worry about thermal noise from a K-3II for astrophotography. That is the camera that Pete_XL uses and gets some phenomenal results.

Agreed. IMO Pete_XL is the reigning champ of K-3II astrophotography. His work is inspirational and was a great influence in my decision to get a K-3II (that, the astrotracer feature and all the other features that make it a great DSLR for general photography). Now if only I can get some time to use it under dark skies!

Re worrying, I'm not exactly worried. With its Sony sensor I have confidence the K-3II has low enough thermal noise to allow great results limited more by my technique than the camera (isn't it always the case?). Nevertheless, it would be nice to see some reliable (i.e., solid, without serious question) data on that. I don't care if Canikons are dismissive of Pentax cameras, either. It's their loss. (The authors of the article I referenced are very experienced astrophotographers but know and trust only Canon and Nikon. Their book - The Backyard Astronomer's Guide - is, by the way, a very good one. It's into its third edition and I've bought them all, even though at this stage there's really nothing in the book that I haven't seen before.)

[maw]

They also don't say what kind of noise they are talking about.

* I just wonder how they can say that Canon is better than Pentax if they haven't tried it or don't know it.... !?! Let us not tell lies without having real data in our hands.

Look, I don't think a thermal noise test exists because all the sensors overheat in relation to temperature and service life.
So it is not a question of trademark.

I chose the K3II because it has some features that interest me more than others, for example the built-in flash is not a prerogative, I looked at GPS,
The image stabilizer, Pixel shift, Astrotracer, Anti-aliasing (AA) filter, and last but not least the K3 has a shutter for 200,000 cycles.
Characteristics that Canon or Nikon do not have in models of equal category. The choice is yours.

I found two face-to-face tests (there are many more) between the K3II and 70D, see if they can help.

1. K3II vs 60D

2. K3II vs 60D

3. K3II vs 60D

[Lew Dite]

I chose the K3II because it has some features that interest me more than others, for example the built-in flash is not a prerogative, I looked at GPS, The image stabilizer, Pixel shift, Astrotracer, Anti-aliasing (AA) filter, and last but not least the K3 has a shutter for 200,000 cycles. Characteristics that Canon or Nikon do not have in models of equal category. The choice is yours.

Yup. Those features and a few others swayed me in favor of the K-3II over something from Canon (e.g., the 80D) or a model from Nikon (e.g., the D7200). I chose the K-3II even though it is officially discontinued, even though it is three years old and is based on a camera (the K-3) that is now five years old. I chose it over the newer K-70 and KP models despite it having worse high-ISO noise levels and despite it having a fixed screen (a tilting one would have been nice) because it just had a better mix of features for my needs. I chose it over the K-1 because the K-1 was twice the price and I didn't want the K-1's extra weight and size (the camera body and the lenses).

[maw]

Those features and a few others swayed me in favor of the K-3II over something from Canon (e.g., the 80D)

I wanted to buy the 80D, because I own the 6D and several Canon L lenses and not, so I was interested to compare the lenses with the APS-C format,
but then I read some reviews not very relevant to my tastes on the 80D. So now I'm going to sell all my Canon kit.

I'm satisfied with the K3II, perhaps more attention to noise would have been welcome.

In this forum from amateurs to amateurs non professionals to professionals, one of the largest and most complete in Italy and beyond, there are many opinions of users and photos taken.

You need to register to see them in real resolution and access all the other functions, highly recommended.

Link: Juzaphoto

Bye, Mario

[Lew Dite]

An update: After my last post in this thread I discovered that one of the authors of The Backyard Astronomer's Guide, Alan Dyer, had done a review of the K-3II for astrophotography that was published in the February 2017 issue of Sky & Telescope, a popular astronomy magazine that has been published since the 1940's. I started a thread about it in the Astrophotography group. You can read Dyer's review (sans photos) here.

To make a long story short, although Dyer found some things he liked about the camera he ultimately panned it as something "of interest first and foremost to Pentax fans wanting to try their hand at astrophotography." In other words, only for beginners who for some reason want to use Pentax. He made several misleading or inaccurate statements in the review concerning its noise performance, such as it having "no worse" noise than a Canon 60Da (while technically true, it should have considerably _less_ noise) and it not using an ISO-invariant sensor like in Nikon and Sony cameras. He doesn't seem to know or care that the sensor on the K-3II is made by Sony and shares all the usual positive characteristics of Sony sensors.