DR disclaimer

[fredjeang]

Hi,

Just a quick disclaimer that I find at least honest to post.

In another recent thread, I was claiming that my past experience in post-production gave me the impression that
the differences in DR between the 35mm and MF was about a maximum of 4 stops.

At the same time, I was aware that it was a subjective evaluation, dictated by both a sensation and also because the room available
in MF files was truly impressive compared to smaller format.

I must add that when I was working in agencies, the 35mm had not reach yet the level of performance they have now,
so those differences where truly exacerbated. (that was the time of the D2).

But I stayed with a desire to verify one day on the field if my subjective sensation was right or wrong.

Well, this has been done now and I can say without reservation that I was wrong.

As I do not own yet an MF back, I had to wait an oportunity to do some comparaisons with a friend, a retired experienced press
photographer that owns a MFD for his pleasure.
Canon 1DsMK3 vs Sinar 31MP if I remember well. (gear specs are not for me)

We saw that the differences where constantly 1 stop and in some circunstances 2 stops (particularly dark subjects or areas).
Isos 100.
In controled light, those differences where almost impossible to detect. Yes, a little more smoothness and transitions in favor of the MF

I realised that the subjective sensation that made me beleive during many time that the DR gap was indeed bigger, is the fact that in
post-production and IF you practice BDSM with the files, the MF stands still while the Canon, well, is more vulnerable.

So after this oportunity to do some informal but informative testing with a friend, I can affirm more reasonably that the DR differences are 1 stops + (ish)
but that the MF file handle much more severe treatment in post-production.

Now to be also fair, I did not have the opportunity to compare with high-end backs like the P65.
 

Cheers.

[ErikKaffehr]

Hi,

Thanks for reporting this. I would say that about one stop difference is what I would expect due to sensor size.

On the other hand I had a short discussion with Mark Dubovoy and he is very affirmative there is a several stop difference between Phase One MFDBs and the Canon 1DsII he tested regarding DR with FULL TEXTURE. He also wrote that Jeff Schewe found a 4 stop difference between the DSLR DR champs (D3X anf 5DII).

So, what we see is that your observations are more in line with some expectations while some real experts see significantly larger benefits with state of the art MFDBs compared with state of the art DSLRs.

It seems that the subject is somewhat controversial. There are some authors on this forum having very good insight in imaging technology from the theoretical viewpoint (Graeme Netress, BJL, ejmartin to name a few) but they would normally not have MFDBs to dissect, so they don't have the opportunity to come up with a theoretical explanation.

Your findings also contradict the view that any MFDB is better than any DSLR.

Best regards
Erik

Hi,

Just a quick disclaimer that I find at least honest to post.

In another recent thread, I was claiming that my past experience in post-production gave me the impression that
the differences in DR between the 35mm and MF was about a maximum of 4 stops.

At the same time, I was aware that it was a subjective evaluation, dictated by both a sensation and also because the room available
in MF files was truly impressive compared to smaller format.

I must add that when I was working in agencies, the 35mm had not reach yet the level of performance they have now,
so those differences where truly exacerbated. (that was the time of the D2).

But I stayed with a desire to verify one day on the field if my subjective sensation was right or wrong.

Well, this has been done now and I can say without reservation that I was wrong.

As I do not own yet an MF back, I'll have to wait an oportunity to do some comparaisons with a friend, a retired experienced press
photographer that owns a MFD for his pleasure.
Canon 1DsMK3 vs Sinar 31MP if I remember well. (gear specs are not for me)

We saw that the differences where constantly 1 stop and in some circunstances 2 stops (particularly dark subjects or areas).
Isos 100.
In controled light, those differences where almost impossible to detect. Yes, a little more smoothness and transitions in favor of the MF

I realised that the subjective sensation that made me beleive during many time that the DR gap was indeed bigger, is the fact that in
post-production and IF you practice BDSM with the files, the MF stands still while the Canon, well, is more vulnerable.

So after this oportunity to do some informal but informative testing with a friend, I can affirm more reasonably that the DR differences are 1 stops + (ish)
but that the MF file handle much more severe treatment in post-production.

Now to be also fair, I did not have the opportunity to compare with high-end backs like the P65.
 

Cheers.

[fredjeang]

Hi,

Thanks for reporting this. I would say that about one stop difference is what I would expect due to sensor size.

On the other hand I had a short discussion with Mark Dubovoy and he is very affirmative there is a several stop difference between Phase One MFDBs and the Canon 1DsII he tested regarding DR with FULL TEXTURE. He also wrote that Jeff Schewe found a 4 stop difference between the DSLR DR champs (D3X anf 5DII).

So, what we see is that your observations are more in line with some expectations while some real experts see significantly larger benefits with state of the art MFDBs compared with state of the art DSLRs.

It seems that the subject is somewhat controversial. There are some authors on this forum having very good insight in imaging technology from the theoretical viewpoint (Graeme Netress, BJL, ejmartin to name a few) but they would normally not have MFDBs to dissect, so they don't have the opportunity to come up with a theoretical explanation.

Your findings also contradict the view that any MFDB is better than any DSLR.

Best regards
Erik

Yes Erik, I also found that 4 stops was about to be right, but I must say that I find that was a very subjective appreciation.
It seems as you point that there is absolutly no consensus on that matter.
I defended the Mark's point of view when that thread showed up at that time and I'm fully aware that Mark is knowledgable.

I hate to say that because I don't like when things are so subjectives, but, could it be that Michael was right when he said: "these are all subjective datas and therefore
can be interpretated differently".

Thank god man did not try to reach the moon with photography science...  

[ErikKaffehr]

Hi,

One point which I think may come into effect is that with DSLRs we normally have a significant sharpening to compensate for the AA-filter. MFDBs need much less sharpening and sharpening enhances noise.

Another issue may be if different processing pipelines are used, like LR on DSL and Capture One on MFDBs.

Just speculating.

Erik

Yes Erik, I also found that 4 stops was about to be right, but I must say that I find that was a very subjective appreciation.
It seems as you point that there is absolutly no consensus on that matter.
I defended the Mark's point of view when that thread showed up at that time and I'm fully aware that Mark is knowledgable.

I hate to say that because I don't like when things are so subjectives, but, could it be that Michael was right when he said: "these are all subjective datas and therefore
can be interpretated differently".

Thank god man did not try to reach the moon with photography science...