No point in using IIQ16bit if you are above base ISO, right?

[narikin]

Just want to check I have this right - I think I read there is no point in using the larger IIQ16bit format if you are working above the base ISO of 100 (for IQ100) - is that right? The gain applied to shadows etc, renders 16bit pointless, I thought.

I shoot fast, usually at 200asa, so IIQ-L has always suited me, but if I'm missing something really, then please advise!
(I can see a very tiny difference at 100asa, but not higher, to my eyes)

Thanks

[BAB]

I would think the color from 16bit files are far better to superior at any ISO setting they will just give you the best possible image capture.

[Wayne Fox]

I would think the color from 16bit files are far better to superior at any ISO setting they will just give you the best possible image capture.

They're both still 16bit but  it's a little unclear exactly what is different about the IIQ L file and the IIQ L 16bit files.

I suppose you might be right that once you push the ISO, any gains from the IIQ L 16bit format would be negated. But I've never seen a conclusive statement from Phase on this.  Sounds like it is specifically designed to work with the 16bit capture of the IQ3 100 ...so maybe not using compromises the quality a little.

From Phase ... "Why IIQ L 16bit?
Both IIQ L and IIQ S are 16bit  le formats when using CCD digital backs (IQ3 80MP/60MP, IQ280/260, IQ180/160/140).
... The IQ3 100MP features a new and Improved CMOS  le format “IIQ L 16bit”. The IIQ L 16bit is, similar to the IIQ L, a completely lossless  le format. It handles output in full 16bit from the IQ3 100MP CMOS sensor which is a new feature introduced with this special CMOS sensor.

While IIQ L 16bit provides the ultimate quality possible from the IQ3 100MP, the file size is also larger and the capture rate slower when compared with the IIQ L or IIQ S."

[ErikKaffehr]

Hi,

You are probably essentially right. It seems that the IQ3100MP has about 15EV of DR per pixel. So you need 15 bits to handle the pixel output. When you go from ISO 100 (nominal) to ISO 200 (nominal) there is a significant drop of DR, around 2/3EV. So DR would be very near to 14 EV at 200ISO. So, 200 ISO may be a border case.

The 15EV of DR responds to engineering DR, while Bill Claff's diagram (enclosed) uses 'Photographic DR' which uses a different criteria.

Engineering DR is essentially maximum well capacity divided by readout noise. That readout noise, in combination with photon statistics provides a dither that eliminates quantisation errors.

Best regards
Erik

Just want to check I have this right - I think I read there is no point in using the larger IIQ16bit format if you are working above the base ISO of 100 (for IQ100) - is that right? The gain applied to shadows etc, renders 16bit pointless, I thought.

I shoot fast, usually at 200asa, so IIQ-L has always suited me, but if I'm missing something really, then please advise!
(I can see a very tiny difference at 100asa, but not higher, to my eyes)

Thanks

[Balafre]

Bravo! Thank you, this was great!