Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
04-15-2012, 09:52 PM   #1
Veteran Member
bwDraco's Avatar

Join Date: Jul 2010
Location: New York
Photos: Albums
Posts: 1,071
Why can't the K-5 JPEG buffer be deeper?

The K-5 has a very deep physical buffer, with a tested worst-case buffer depth of 20 shots (RAW+JPEG 16M Premium, ISO 51200, 1/8000s). Under typical shooting conditions, the K-5 can shoot 30 JPEG/21 RAW@7 fps. Why can't the JPEG buffer be much deeper as with the Canon EOS 7D (94 JPEG/15 RAW@8 fps) and Nikon D300S (44 JPEG/18 RAW@7 fps)?

The K-r has a much shallower worst-case buffer of 6 1/2 shots (6 shots, followed by a slightly delayed 7th shot), but a JPEG buffer of at least 25 shots in typical conditions. What would explain this anomaly?

I'm using a 16GB SanDisk Extreme 30MB/s card.

I know this topic has been discussed several times before, but can anyone give a satisfactory technical reason for this behavior?

--DragonLord


Last edited by bwDraco; 04-15-2012 at 10:11 PM.
04-15-2012, 10:07 PM   #2
hcc
Pentaxian
hcc's Avatar

Join Date: Nov 2009
Location: Brisbane, Australia
Photos: Gallery | Albums
Posts: 3,995
The buffer size is relevant if you use large file sizes, if your card is slow and/or if you use some in-camera PP.

I use a K-7 and I can shoot 14 Mp JPEG [**] continuously. I have taken a number of long sequences, incl. for over 60 s! I use a Sandisk Extreme III class 10, no in-camera PP and [**] JPEG image quality.

While a larger buffer size might be of interest to a few, the K-7/K-5 has already a relatively large buffer and it is easy to get long continuous series of shots with the correct settings.

Food for thought...
04-15-2012, 10:12 PM   #3
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 51,558
I guess the limit is there because there's a cap on how quickly in-camera processing can be performed. I'm looking forward to seeing how the Prime M will do in the new high-end DSLR, as obviously they won't constrain the buffer to just 6 frames in the new camera as they did in the K-01.

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:
04-15-2012, 10:14 PM   #4
Veteran Member
bwDraco's Avatar

Join Date: Jul 2010
Location: New York
Photos: Albums
Posts: 1,071
Original Poster
I can't quite explain it, but it seems that the amount of time needed to write a single shot to the card can vary greatly even if the shot is nearly identical. I don't know what can account for this variation, but it shouldn't happen. A given JPEG shot can take anywhere from about 0.3 to 1.1 seconds to write--and this is with the same settings and composition. Could this be part of the problem?

--DragonLord

04-15-2012, 10:21 PM   #5
Veteran Member
bwDraco's Avatar

Join Date: Jul 2010
Location: New York
Photos: Albums
Posts: 1,071
Original Poster
QuoteOriginally posted by Adam Quote
I guess the limit is there because there's a cap on how quickly in-camera processing can be performed. I'm looking forward to seeing how the Prime M will do in the new high-end DSLR, as obviously they won't constrain the buffer to just 6 frames in the new camera as they did in the K-01.
I've got a feeling that the processor in the next-generation DSLR will be called PRIME III, as the M designation is probably related to the mirrorless nature of the K-01. We can reasonably expect this new processor to inherit the new technologies in the PRIME M processor (including UHS-I support), as well as further enhancements in performance. Given that the latest SD cards are closely approaching CF card write speeds (90+ MB/s), we should not have to use CF cards to get the sort of performance we see in competing C&N cameras.

--DragonLord
04-15-2012, 10:26 PM   #6
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 51,558
QuoteOriginally posted by DragonLord Quote
I've got a feeling that the processor in the next-generation DSLR will be called PRIME III, as the M designation is probably related to the mirrorless nature of the K-01. We can reasonably expect this new processor to inherit the new technologies in the PRIME M processor (including UHS-I support), as well as further enhancements in performance. Given that the latest SD cards are closely approaching CF card write speeds (90+ MB/s), we should not have to use CF cards to get the sort of performance we see in competing C&N cameras.
Given how many components Pentax recycles going from one camera to the next, I beg to differ. But I guess we will have to wait and see- maybe Pentax will surprise us (or stick two processors in there like Canikon are doing)!

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:
04-16-2012, 12:04 AM   #7
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
The K-01 only manage 1fps with RAW and i don't see them magically 6 to 7 fold the performance so lets hope they don't use the same processor....

About the JPEG buffer, might have something to do with the algorithm Pentax use, you need to compare the quality of the files as well.

04-16-2012, 12:11 AM   #8
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 51,558
Given that the K-01 is faster overall (during playback and image processing) and supports 60FPS video, I think it's safe to say that the processor is powerful enough but the firmware is designed to limit the framerate. As stupid as that sounds, I can see them saving higher specs for more expensive bodies...

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:
04-16-2012, 12:12 AM   #9
Pentaxian




Join Date: Apr 2007
Location: Sweden
Posts: 2,106
Maybe the raw to jpeg conversion is a bit slow, so that in reality the buffer will hold mostly raws even when shooting jpeg. Just a thought.
04-16-2012, 12:38 AM   #10
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
QuoteOriginally posted by Adam Quote
Given that the K-01 is faster overall (during playback and image processing) and supports 60FPS video, I think it's safe to say that the processor is powerful enough but the firmware is designed to limit the framerate. As stupid as that sounds, I can see them saving higher specs for more expensive bodies...
But why on earth give the K-01 only 1FPS?
All of the direct competition are almost 3 times as fast...
The new Milbeaut can process 7 frames of 16mp continuously till the SD card is full so they have crippled it a lot, the processor should be working around 1/10th or even lower then it can do.... highly doubtful. I also think they use a different processor for it, it could be the one also used in the Q maybe.

Video has little to do with it, the processor in the K5 does not actually support video on the hardware level so therefore it's not comparable.
And it's 30 frames not 60 in 1080p.

Last edited by Anvh; 04-16-2012 at 12:49 AM.
04-16-2012, 03:46 AM   #11
Veteran Member
bwDraco's Avatar

Join Date: Jul 2010
Location: New York
Photos: Albums
Posts: 1,071
Original Poster
QuoteOriginally posted by Anvh Quote
Video has little to do with it, the processor in the K5 does not actually support video on the hardware level so therefore it's not comparable. And it's 30 frames not 60 in 1080p.
Video on the K-5 is limited to 25 fps at 1080p, and 30 fps at 720p and 480p.

-----

Why would the K-5 buffer for JPEG be not that much deeper than that of the K-r, despite having a far deeper buffer for RAW? The K-r can shoot 25+ JPEG frames with a buffer not capable of holding more than 6-10 RAW+JPEG frames, which makes it appear that the K-5 JPEG buffering is less efficient than that of the K-r.

Perhaps the 14-bit image processing pipeline in the K-5, compared to 12-bit in the K-r, would explain this discrepancy?

--DragonLord

Last edited by bwDraco; 04-16-2012 at 03:53 AM.
04-16-2012, 07:24 AM   #12
Veteran Member




Join Date: Jun 2009
Posts: 11,913
Shoot lower ISO for optimal JPG in the buffer. The testing is usually performed at about ISO 100, you will find. High ISO means slower fps.

Real world performance (eg in sports shooting) will vary wildly as the ISO jumps around as the scene and exposure requirements change - for any Canon, Nikon or Pentax.
04-16-2012, 10:30 AM   #13
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
I know the K5 support video but the processor doesn't from the factory, that's probably why MPEG is used and not another file format.
So the K-01 processor with H.264 support baked in so much more optimized for doing video.

For JPEG difference i think you need to look into the quality of both files, the Kr could use a more aggressive and faster way to convert the files.
04-16-2012, 09:21 PM   #14
Veteran Member




Join Date: Sep 2006
Location: Kaunas
Posts: 1,458
QuoteOriginally posted by DragonLord Quote
The K-5 has a very deep physical buffer, with a tested worst-case buffer depth of 20 shots (RAW+JPEG 16M Premium, ISO 51200, 1/8000s). Under typical shooting conditions, the K-5 can shoot 30 JPEG/21 RAW@7 fps. Why can't the JPEG buffer be much deeper as with the Canon EOS 7D (94 JPEG/15 RAW@8 fps) and Nikon D300S (44 JPEG/18 RAW@7 fps)?

The K-r has a much shallower worst-case buffer of 6 1/2 shots (6 shots, followed by a slightly delayed 7th shot), but a JPEG buffer of at least 25 shots in typical conditions. What would explain this anomaly?

I'm using a 16GB SanDisk Extreme 30MB/s card.

I know this topic has been discussed several times before, but can anyone give a satisfactory technical reason for this behavior?
Just a guess.

Well, maybe Pentax image processing is:
Sensor -> Buffer (RAW data) -> Prime II -> SD card

while Canon is:
Sensor -> Dual Digic 4 -> Buffer (processed jpeg/RAW file) -> CF card

In other words: Canon 7D processors are powerful enough to process images on the fly and buffer is used to store processed files before they are written to CF card.
Pentax K-5 Prime II can't process images on the fly, so they are queued in buffer for processing.
So, in Canon 7D case not only jpegs need less data to be transferred into CF card, but they take up less space in buffer, while in K-5 case difference between RAW and jpegs capacity is only because jpegs are written to SD card quicker (smaller size).
04-16-2012, 11:06 PM   #15
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
Pentax goes as followed.
Sensor > RAM > Processor > RAM > SD card

The actual processor for the image is different then the one that writes the file to the SD card, they are separate.

I think everyone uses this model because no processor can keep up the frame rate and if the processor is still busy with a photo it would mean the photo in the "pipeline" will get lost. The sensor needs to write it to somewhere before it can take another photo.
Speed of the RAM is probably 3200MB/s so clearly not the bottleneck in this arrangement.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
buffer, camera, conditions, dslr, fps, jpeg, k-5, k-5 ii, k-5 iis, k5, pentax k-5, shots
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Would a better lens give me deeper DOF? reggieperrin Troubleshooting and Beginner Help 19 05-22-2011 08:44 PM
k-7 4 shot buffer? cwood Pentax DSLR Discussion 5 02-12-2011 10:18 PM
Larger buffer might actually be HW? Gimbal Pentax DSLR Discussion 38 11-02-2010 06:21 AM
K5 Buffer BrendanPK Pentax K-5 & K-5 II 133 10-30-2010 01:09 AM
K-7 40 jpeg buffer enough? Pentaxor Pentax DSLR Discussion 5 06-28-2009 01:28 PM



All times are GMT -7. The time now is 10:47 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top