iTTL Nikon Protocol – Metz SCA300 interface part-3

back to blog       next post

Metz SCA 3000 interface for Metz45 – part 3

This is the last round of the implementation of Nikon protocol to manage Metz 45 CL-4.  In this document I will focus on the finalization of the implementation of this protocol connecting Metz 45 CL-3 or CL4 with Nikon DSLR.

 

After some adjustment of the pre-flash duration and introduction of power duration calculation taking into account: ISO, requested power, Expo compensation, I obtain a usable tool ready for new event. With GN 45 at ISO 100 / 35mm FX (meter) it give a comfortable light ambiance especially using indirect/bounce flash.

 

Enhancement done

Read Analog input

Implementation of non-blocking analog read to avoid losing command or Sync event. AnalogRead function was split in two parts: start conversion and read result. It permits to do conversion in background, and CPU to process real-time event “in time”.

Exposure calculation

Management of ISO setting. The flash have to calculate exact light quantity taking in consideration the requested power, ISO setting, EV adjustment from DSLR.

Pre-flash level adjustment

Adjust power level of the two different pre-flash and adjustment of the overall exposure.

Report version 1.000 to DSL

Report software version to DSLR in A1 command

Red eyes reduction

Implement Red eyes reduction as a non-blocking functionality

Overheating security

Implement over heating security (in case of usage of Quantum Turbo and external power pack) only in TTL mode - max 10 shot at 1/1 in 20sec then wait.

use only preflash2

force the system to use always the second pre-flash

 

Limitation

With original 45CL3 or 4 flash gun, due to Thyristor circuit design, TTL is only usable with AEL (FL mode).

  - pre-flash to memorize the exposure

  - take the picture.

IGBT is the solution to remove this limitation.

 

Residual problem 

During test, I identify some problem like shot missing or over exposed. This is due to bad transmission (ISO too high..., cmd not received) especially when I activate the trace for debugging. Works well with no debugging (seems that we are not able to process all events in real-time).

 

Last version of source code

 Arduino software to manage i-TTL protocol and Metz 45 CL-3 or CL-4 flash        45-iTTL_V1.ino

 

Exposure calculation

Working on the Exposure adjustment in i-TTL mode I discover that ISO is managed by flash. The power command take into account: selected aperture and reflectance of the subject. Flash must calculate strobe duration based on required power (0xD3 command) ISO setting and Exposure adjustment sent by the DSLR.

Required power send by DSLR is based on pre-flash result taking into account the reflectance of the subject, max aperture of the lens and selected aperture.

 

Pre-flash

There are two pre-flash with two different level (modification of the first report – second pre-flash is not 30% power but 30% more than preflash1)

Pre-flash 1, the first is always used. It is mainly for prime lens, < f 3.5) and certainly for short distance.

Pre-flash 2, could be used if result of pre-flash one is not sufficient. Pre-flash 2 send near 30% more light than Preflash1. It is used when reflected light is not sufficient (Zoom with f4.. f5.6  (and smaller) full aperture characteristics or indirect /bounce flash

It means that the system is adaptive, keeping light level in the range of the DSLR light metering system. It permits:

-          to use prime or standard lenses

-          to high/low reflective subject or short / long distance with subject.

Side effect: the system could produce one or two pre-flash but nor really a problem if we stay in the Nikon light system environment.

 

Overheating security

I implement overheating security to block flash if you do 10 shot at full power (1/1) in 20sec. It is very easy with Quantum Turbo battery. In this case I report flash not ready. I have to investigate a little more on this area because DSLR do not wait flash ready and shot without flash. As it is an extreme situation it is not so problematic at the moment.

 

 

Document version

V1.1 – 15 May 2020 – © Hervé QUEVAL – herve.queval@free.fr