GH1: 8bit sensor vs 7D: 14bit sensor

M

mo sizzle

Active member
I noticed that the gh1's m4/3 sensor is 8 bit and the 7d aps-c sensor is 14 bit. I know they both have 4:2:0 color sampling, but wouldn't the 6bit difference offer a huge advantage to detecting accurate color for the 7d?
 
Look at the footage from both. Technical jargon sometimes is easy to get caught up in. Don't worry about it. Does the GH1 produce footage that you like? Does it compare to 7D/T2i?

I'd say it does. Looks great to me. I don't care if it's 1bit... it looks fantastic.
 
I agree, I've owned the gh1 since it was released, and have been using it and love it. But the difference in color bit could mean much better compositing, color correction, etc. Just wondering.
 
But any potential advantages would be lost (or tremendously reduced) when the footage is encoded with the hardware encoder to 8-bit h.264 , unless you were able to access the raw stream before compression
 
mo sizzle said:
well some people edit in prores 4444 which is 12 bit i believe
Click to expand...

Well if you start with a 8-bit 4:2:0 source, the values are interpolated.

It's not going to be as good as a true 10-bit or 12-bit workflow preserved from start to finish

The "weak link" is the 8-bit 4:2:0 h.264 compression that is being used as a source
 
Ya.. the only way to get full 10bit would be to capture off of the HDMI and bypass the compression. I doubt it would be worth it unless you really needed 10bit for composting/greenscreen work.
 
oh alright, didnt know that. but the 7d still samples color at 16 bit before it gets compressed down to 8 bit right?
 
ATL Media Group said:
Ya.. the only way to get full 10bit would be to capture off of the HDMI and bypass the compression. I doubt it would be worth it unless you really needed 10bit for composting/greenscreen work.
Click to expand...

Correct me if I'm wrong (I know someone will!), but isn't this a moot point? I don't think the 7D outputs an uncompressed signal via HDMI. Or does it? I think it outputs a lower rez signal that is often used for monitoring. Still useful...and beats the GH1 which outputs nothing during record. But again, I might be wrong. I don't have a 7D.
 
I dunno if it does.. Ya it would be a moot point (I like that ya got it right... drives me nuts when folks say "mute point") if it doesn't send out an uncompressed signal. I assumed it did as most cameras with HDMI out send the signal out before processing.
 
i bet the 10-bit data is used for stills which is a big part of the reason the canon's beat the gh1 in photos (apart from megapixel diff too!).
 
averan said:
i bet the 10-bit data is used for stills which is a big part of the reason the canon's beat the gh1 in photos (apart from megapixel diff too!).
Click to expand...

Well, not necessarily - especially when you compare the 7d to the GH1. From DxOMark's review of the GH1:

"The GH1 even outperforms a hypothetical four-thirds sensor cut from a (wider) Canon EOS 7D sensor (with the same pixel size). With its 50% larger sensor surface (corresponding to 9 DxOMark Sensor points), one would expect the 7D to have outperformed the GH1 by 9 points. Instead, it achieves a DxOMark score of 64.9 — only slightly more than 1 point above the GH1’s score of 63.6."

Comparison with a similar pixel pitch sensor: Canon EOS 7D

The Canon EOS 7D uses pixels of 4.29 µm vs 4.31 µm for both the GH1 and G1. To compare the pixel performance (and not the whole sensor), we can compare the performances in “screen” mode.
SNR 18%

Obviously the Canon 7D’s performance is similar to the G1’s at ISO settings up to 800, and similar to the GH1’s at ISO settings from 3200. The GH1 outperforms the 7D for ISO settings up to 800.

Dynamic Range

The 7D and GH1 have similar performances, and both outperform the G1 by more than 1 stop.


Color sensitivity

The GH1 outperforms both the G1 and 7D at ISO settings up to 800; over ISO 800, GH1 and 7D performances are similar.
 
cowpunk52 said:
Well, not necessarily - especially when you compare the 7d to the GH1. From DxOMark's review of the GH1:[/B]
.
Click to expand...



Good material. The GH1 is always getting slagged. Even if tests come out like these, it seems not prevent/reduce negative conversation about it. Now if the bit rate was better, then the chatter on the GH1 would be less negative though. I will be looking forward to seeing how the GH1 does in the much talked about DSLR tests vs 35mm.



Michael Pappas
http://www.pbase.com/Arrfilms
http://www.PappasArts.com
http://SumeriusChronicles.com/
http://TheOrbisTerrarum.com/
http://TheOrbisTerrarum.blogspot.com/
http://SumeriusChronicles.blogspot.com/
http://www.facebook.com/PappasArts
 
What I see as a problem in DSLR cameras isn't necessarily bit depth. It's how we resolve what we do have. While having greater bit resolution would definitely give you an advantage in actual frame resolution, I think the mindset behind the designs is at odds with what we really need.

I have noticed that my GH1, like most other cameras, can have blown out highlights and crushed black shadows in the same frame yet seem both too bright in lighted areas and too dark in partially dark areas. It seems to be a byproduct of the designers trying to get too much contrast into the frame since they know that contrast makes things look "pretty", at the expense of ruining the mid-range contrasts that our eyes can pick out very easily, like details in shadows/highlights and such.

I did a simple test where I found a very contrast-y scene with a very bright highlight and very dark shadow. I opened up the iris until most of the frame was blown out yet the shadow remained too dark to see any detail. I did the opposite, and stopped down until nothing but the highlight remained. It still looked completely blown out yet I couldn't see anything else in the frame. Histogram didn't show a blown white either.

Thoughts?
 
Measurebate all you want, but really it comes down to you pressing that little shutter release button on the top handle and seeing which camera gives you an image you are happy with, and an image you can manipulate in post and not have fall apart (if you want to do heavy post processing). This is for still work.

For video why you would even deal with the sensors' specs is kind of beyond me. The internal processing guts the image so much to get all that information (24 frams per second) down that I would be highly impressed if anyone was actually able to tell.
 
William_Robinette said:
Measurebate all you want, but really it comes down to you pressing that little shutter release button on the top handle and seeing which camera gives you an image you are happy with, and an image you can manipulate in post and not have fall apart (if you want to do heavy post processing). This is for still work.

For video why you would even deal with the sensors' specs is kind of beyond me. The internal processing guts the image so much to get all that information (24 frams per second) down that I would be highly impressed if anyone was actually able to tell.
Click to expand...

My post wasn't meant to "measurebate" anything - but good measurements do play a vital role in debunking uninformed speculation, i.e. that the Canon 7d performs better than a GH1 for stills because of more bit depth.

I wasn't referring to video capabilities of either camera.
 
Svart said:
What I see as a problem in DSLR cameras isn't necessarily bit depth. It's how we resolve what we do have. While having greater bit resolution would definitely give you an advantage in actual frame resolution, I think the mindset behind the designs is at odds with what we really need.

I have noticed that my GH1, like most other cameras, can have blown out highlights and crushed black shadows in the same frame yet seem both too bright in lighted areas and too dark in partially dark areas. It seems to be a byproduct of the designers trying to get too much contrast into the frame since they know that contrast makes things look "pretty", at the expense of ruining the mid-range contrasts that our eyes can pick out very easily, like details in shadows/highlights and such.

I did a simple test where I found a very contrast-y scene with a very bright highlight and very dark shadow. I opened up the iris until most of the frame was blown out yet the shadow remained too dark to see any detail. I did the opposite, and stopped down until nothing but the highlight remained. It still looked completely blown out yet I couldn't see anything else in the frame. Histogram didn't show a blown white either.

Thoughts?
Click to expand...

This is a symptom of digital sensors, and has been for a long time. In general, they don't have as nice of a roll-off of highlights & shadows compared to, say, film. Take a look at the comparisons between a D300 and film here:

http://www.twinlenslife.com/2009/05/digital-vs-film-real-deal-nikon-d300-vs.html

Before anybody gets up in arms about, this comparison has nothing to do with sensor or frame size. Just look at how each handles light, shadow, detail & exposure. You'll have a better understanding of the limits of digital sensors when dealing with wide contrast.

As for your GH1 shots, try setting the contrast down as low as it will go in "smooth" mode to get the most dynamic range you can while keeping the roll-off from being too harsh. Also, by shooting in RAW mode you can eek out a little more shadow and highlight detail in post. The abrupt roll-off of the GH1 of over and underexposure is a well-documented issue of this camera (and many other digital cameras).

edit: I've also noticed that by reducing in-camera sharpness, you reduce contrast as well (to a smaller degree)
 
Last edited:
I think your wrong?

I think your wrong?

The 14 bits you are referring to is the raw sampling, I think the gh1 are actually 12 bit raw, but I'd need to check, apart from that the only difference really is the sampling of the video which is very high on canon and only 17mps on gh1. Colour depth in video will be exactly the same. The gh1 is not an 8bit sensor, this is technically wrong?
 
You must log in or register to reply here.
Back
Top